Вы находитесь на странице: 1из 64

Adolescent Attachment

Running Head: ADOLESCENT ATTACHMENT

Adolescent Attachment, Peer Relationships, and School Success:


Predictor, Mediator, and Moderator Relations
Jill Carlivati
Distinguished Majors Thesis
University of Virginia
April, 2001

Advisor: Dr. Joseph P. Allen


Second Reader: Dr. Nancy S. Weinfield

Adolescent Attachment

Acknowledgements
Many individuals helped make the writing of my thesis a rewarding learning
experience. Through their unselfish support, I met each project challenge confidently as I
gained new insights into both adolescent development and research methodology.
I thank Dr. Joseph P. Allen not only for his ongoing guidance with the project, but
also for his leadership with the Virginia Study of Teens and Families, which provided the
framework for my study. My frequent discussions with Dr. Allen on underlying theory
and project design proved invaluable. His direction and support provided clarity,
discipline, and meaning to my work.
Similarly, I express deep appreciation to Dr. Nancy S. Weinfield, not only for
serving as my second reader, but also for her instruction on the nature of attachment. Her
expertise in this area, as well as her willingness to share this knowledge, enhanced my
understanding of adolescent attachment and offered me new insights.
Additionally, I extend a sincere thank-you to Heather Tencer for serving as my
graduate advisor. Her assistance with data analysis and the observations she made on
early versions of the paper were particularly valuable.
Other members of Dr. Allens lab, particularly Penny Marsh, Felicia Hall, and
Maryfrances Porter were also very helpful to me as I pursued my work. I sincerely
appreciate their support with SAS programming and school data analysis.
Finally, to my family, thank you for first teaching me the importance of parentchild relationships.

Adolescent Attachment

Abstract
Because of the influences of school success on diverse areas of development and the
empirically suggested connections among adolescent attachment, peer relations, and
academic factors, the relationships among these constructs were investigated. A sample
of 176 ninth and tenth graders at risk for academic difficulties participated in the study.
Independent predictive links between attachment and school factors, attachment and peer
factors, and peer and school factors were analyzed. The possibility of peer factors
serving as a mediator of the relation between attachment and school factors was
considered. Finally, interactions between attachment and peer factors in predicting
school satisfaction and performance were examined. Independent links among the three
constructs, as well as some mediated and moderated effects of peer relationships, were
supported. Results suggest that a model considering both attachment and peer
relationship factors may best explain teenagers satisfaction with school and performance
in the academic environment.

Adolescent Attachment

Adolescent Attachment, Peer Relationships, and School Success:


Predictor, Mediator, and Moderator Relations
Current research suggests that school performance correlates with quality of ones
interpersonal relationships and additional developmental outcomes. Academic failure, as
assessed by the frequency of problems with peers and difficulty with school work, has a
great impact on relationships in the home, as students who have had a bad day at school
are more likely to display aversive behavior towards their parents that evening (Repetti,
1996). Performance in the classroom can influence self-perceptions, as teacher-rated
work and social patterns predict self-esteem, while grades in a particular subject can
influence students self-esteem relating to that given subject (Hoge, Smit, & Hanson,
1990).
In addition to relationships in the home and ones self-view, experiences in the
school environment can affect ones success in society. The absence of a completed high
school or college education can affect job prospects. Approximately half of the American
population does not have a college education, and national support, especially in the job
market, remains limited for those without high school diplomas. Furthermore, people
without college degrees have significantly lower income than college graduates, and this
monetary disparity is increasing; on average, a high school graduate of the late 1980s has
only one third of the adjusted income of a high school graduate of the 1970s. These
differences in income affect many facets of individuals lives, as adolescents who have
less optimistic job prospects due to a lack of schooling are likely to compensate for their
lessened income by delaying marriage and families (William T. Grant Foundation, 1988).

Adolescent Attachment

Poor school performance also has consequences on physical health. Academic


achievement has been shown to influence the association between the parent-child
relationship and a teenagers involvement in a pregnancy (Scaramell, Conger, Simons, &
Whitbeck, 1998). Among females, lower expectations for academic success are found to
influence the progression from experimental to regular cigarette smoking (Chassin,
Presson, Sherman, Montello, & McGrew, 1986).
Therefore, many aspects of development during the adolescent years are greatly
shaped by success in the school environment. As teenagers progress through middle and
high school, their lives are influenced by both their perceptions of their school
performance and their actual academic achievement. When an adolescent receives lows
grades, is uncommitted to class work, reacts negatively to teachers, and interacts poorly
with peers, a large portion of their day is spent in an environment in which they feel
ambivalent at best, and long lasting influences on development are likely.
Because of the influences of school performance on psychological and biological
development and the influences of school attrition on job and family prospects, it is
important to investigate predictors of adolescents negative scholastic performance.
Hopefully, such work will allow for the creation of intervention measures that will
prevent teenagers from following a path to school failure and general maladjustment
during their teenage years and beyond.
A report by the William T. Grant Foundation (1988) suggests that programs
designed to produce school success will likely fail unless implemented along with family
and community measures. Without the support of those around a child, motivation to
perform well in school, as well as positive reinforcement to do so, may be absent. In

Adolescent Attachment

particular, the adult-child, or more specifically the parent-child, relationship must be


targeted to successfully advance academically at-risk teenagers (William T. Grant
Foundation, 1988). One aspect of the parent-child relationship that has a great impact on
numerous areas of childrens functioning, which would therefore be of interest when
studying predictors of poor school adjustment, is attachment. That is, as attachment has
been shown to relate to peer relationships and peer competence across developmental
periods (Elicker, Englund, & Sroufe, 1992), popularity (DeMulder, Denham, Schmidt, &
Mitchell, 2000), aggression (Cohn, 1990; DeMulder et al., 2000), self-esteem (Armsden
& Greenberg, 1987), and the emergence of depression during adolescence (Nada Raja,
McGee, & Stanton, 1992), there is the possibility that attachment may also be related to
academic functioning.
Attachment is a class of social behavior. Bowlby (1969) characterized attachment
early in development by a childs proximity seeking behavior. At about four months of
age, a secure infant will begin to respond differentially towards adults, preferring the
mother, who serves as the attachment figure. Beginning around age three, a securely
attached child may begin to feel comfortable with teachers or relatives, who serve as
subordinate attachment figures, as the secure child is confident that he or she may
resume contact with the mother as necessary (Bowlby, 1969). As they grow older,
securely attached children continue to explore the world under the care of responsive,
sensitive parents, from whom they seek comfort when distressed (Bowlby, 1969).
Bowlbys theory on attachment is concurrent with the view of attachment as an
organizational construct, an assertion that early learning experiences of the infant within

Adolescent Attachment

the context of the parent-caregiver relationship will continue to influence development


outside of this specific relationship (Sroufe & Waters, 1977).
Variations in psychological adjustment which can be attributed to differences in
attachment security may be reflections of the varying approaches taken by individuals of
different attachments when faced with distress cues (Kobak & Cole, 1994). Children
who have insecure attachment relationships with their caregivers feel much less
confidence in the availability and helpfulness of their parents when in distress (Bowlby,
1969). Insecurely attached children will perceive others as unreliable or unnecessary
sources of comfort (Bowlby, 1973). These children often have problems in diverse areas
of development, exhibiting both emotional distress and often disturbances in personality,
both of which are less likely to be found in securely attached children due to their more
positive representations of interpersonal interactions (Bowlby, 1977).
Furthermore, into the teenage years attachment continues to influence individuals
perceptions of others and their actual relationships. Insecure attachment organizations
during the teenage years may result in difficulties in interactions with others, as well as
perceptions of others, resulting in social problems throughout later life (Allen et al., 1998;
Cassidy, Kirsh, Scolton, & Parke, 1996; Dodge, 1993; Slough & Greenberg, 1990).
Insecure adolescents lack coherence in thinking about attachment-related experiences and
have difficulty processing elements of their peer relationships (Allen & Land, 1999).
Hence, attachment theory leads to the conclusion that there is a causal relationship
between formative experiences with parents and ones ability to form and maintain later
positive relationships with others (Bowlby, 1989), an assertion supported by current
research.

Adolescent Attachment

Armsden and Greenberg (1987) state that attachment relationships may buffer
children from harmful outcomes such as anxiety, depression, and emotional distress, all
of which may occur during times of life turmoil. This buffering hypothesis is especially
relevant to attachment during the teenage years, as early adolescence is inherently a
period of transition (Armsden & Greenberg, 1987). Teenagers moodiness, the often
changing nature of adolescent relationships, and teenagers strives for independence from
their parents may make adolescence a period of continual activation of the attachment
system (Allen & Land, 1999). Therefore, current research suggests that both attachment
organization and current attachment relationships continue to be of great importance
during adolescence, even though this notion has been somewhat counterbalanced by
Bowlbys (1969) earlier view of adolescent attachment, which argues that during
adolescence, parent-child attachment often becomes weaker.
Allen and Land (1999) suggest that although teenagers behavior often appears to
distance adolescents from attachment figures, this observation can be explained by taking
into account developmental changes of the period as well as the acknowledgement that
autonomy seeking behaviors may hinder the recognition of attachment relationships. As
adolescence is a period of change, many teenagers at this stage of life will begin to take
on the role of a caregiver for others, in some cases peers, within attachment relationships
(Bowlby, 1969). This serves as evidence that the individuals role in the attachment
relationship during adolescence often extends beyond a receiver of parental care (Allen &
Land, 1999). In fact it may be necessary for teenagers to seek emotional distance from
their caregivers, allowing their newly attained cognitive processes the freedom to analyze
their previous attachment relationships and to resolve prior relationship issues before

Adolescent Attachment

these severely hinder peer relationships (Allen & Land, 1999). However, it is important
to note that the attachment relationship formed with a parent or caregiver is unique, and
cannot be substituted by another close relationship, such as that with a sibling (East &
Rook, 1992) or a peer. Hence, it would not be surprising that because of the continued
role of attachment in development, attachment security during the teenage years may
predict various peer and school outcomes, two of the major developmental focuses of
adolescence.
It is important to note that attachment during adolescence cannot be defined in the
same way as Bowlbys construct of attachment during early childhood. Following
childhood, security appears to shift from a quality of an attachment relationship to a
description of ones attachment organization, a reflection of ones ease with the
integration of attachment information gathered from the individuals relationship history
(Main, Kaplan, & Cassidy, 1985). Secure adolescents have secure internal working
models, which are used to make inferences about social relationships. Internal working
models, which are the selfs initially conscious and later unconscious mental
representational models of the self and others, are greatly influenced by experiences with
ones attachment figures. Early models of attachment relationships take into account past
experiences a child has had with caregivers, and reflect their caregivers quality of
response (Main et al., 1985). Although internal working models are present at a young
age and are important in an infants perception of their role in their relationship with a
caregiver (Allen & Land, 1999), some argue that it is only as cognitive development
proceeds that children of different security backgrounds will exhibit different internal
working models (Main et al., 1985).

Adolescent Attachment 10
With the onset of adolescence, individuals have reached a level of cognitive
development advanced enough that they may generalize from their past relationships and
use abstract thinking processes to allow one model of attachment organization to emerge
from their past experiences (Allen & Land, 1999). The internal working model comes to
reflect an adolescents state of mind regarding attachment and interpersonal relationships,
presumably allowing ones expectations of the availability of others to remain fairly
constant throughout life (Bowlby, 1973). Hence, adolescent attachment security becomes
a reflection of a teenagers integrated view of relationships, rather than simply a
characteristic of the adolescents relationship with a parent.
It follows that because relationships, both with peers and teachers, should
contribute to adolescents satisfaction with and performance in school, attachment
security may be important in explaining scholastic outcomes during the teenage years. In
fact, there are strong correlations of attachment with both internalizing and externalizing
problems throughout development, especially during adolescence (Allen, Moore,
Kuperminc, & Bell, 1998), and these problems not surprisingly would affect ones
success and happiness in school. Therefore, it is logical to look to attachment both to
predict teenagers social adjustment in the school environment and to interact with peer
relationship factors to produce varying school outcomes. Consequently, attachment and
peer relationships in the classroom may together explain adolescents perceived and
actual academic problems, as insecure teenagers may be overly sensitive to interpersonal
difficulties, neglecting their interactions with others and their learning in the scholastic
setting.

Adolescent Attachment 11
Before it is possible to analyze the connections between attachment and school
performance in adolescence, it is necessary to further clarify the construct of adolescent
attachment and how this construct affects teenagers in the classroom environment. As
adolescents transition to the formal operations stage of cognitive development,
individuals may begin to analyze all of their relationships (for example, those with peers
or teachers) and may even alter their internal working models (Allen & Land, 1999;
Bowlby, 1988; Main et al., 1985). Allen and Land (1999) have offered support to this
assertion by explaining that adolescents newly acquired ability to consider their prior
attachment relationships from the perspective of the caregiver may serve to allow them to
later take on this role in their relationships with peers. Furthermore, as adolescents begin
to analyze past relationships, their strategy employed in approaching attachment
memories, emotions, and relationships becomes crucially important (Allen & Land,
1999). Attachment during the teenage years then becomes ones organization and
processing of attachment-related thoughts, feelings, and emotions, rather than a specific
categorization of a given relationship (for example, having a secure attachment
relationship with ones teacher) as in infancy. However, attachment during adolescence
continues to relate to the manner in which individuals approach social environments such
as the classroom, as the individuals attachment organization still affects ones
relationships with others.
Although much support is offered for a change in the definition of attachment
beginning in adolescence, this is a topic of debate. Bowlby (1969) has stipulated that the
major transition of this period may be a change in the people to whom attachment
behavior is directed. For instance, teenagers may be attached to their friends and their

Adolescent Attachment 12
teachers. That is, Bowlby (1989) has argued that people of all ages are best adjusted
when they have confidence in the availability of an attachment figure. Individuals should
continue to seek out such figures throughout life during times of distress (Allen & Land,
1999). A teenager with an insecure attachment will gravitate towards insecure peers,
likely enhancing the negative effects of such attachments for both individuals. This effect
may occur solely in the immediate context of the relationship (for example, negative
functioning in school if this is the environment where the peers interact) as would be
expected by Bowlbys attachment theory, or may have more overarching effects for all of
a teenagers relationships.
Recent research now suggests, however, that the early view of adolescent
attachment as differing from attachment in early childhood only in whom the behavior is
directed towards is incorrect. Kobak and Sceerys (1988) study, which relied on
adolescents newly acquired ability to integrate attachment histories to form a working
model of the parent-child relationship, found that there is a coherence of attachment
organization during late adolescence. Kobak and Sceery (1988) concluded that distress
activates the attachment system, with the relative coherence of working models
influencing the perceptions of others during distressing events. The higher levels of
familial support of secure adolescents should result in perceptions of availability of
attachment figures during times of distress, whereas the higher levels of anxiety found in
preoccupied individuals should lead to unhealthy dependency in relationships and the
inability to calm distress (Kobak & Sceery, 1988). Therefore, this research suggests that
adolescent attachment organization differs from attachment during infancy, as ones

Adolescent Attachment 13
personal organization of attachment relationships in general becomes the crucial defining
aspect of attachment.
It is possible that attachment organization is only meaningful in the context of an
attachment relationship (Allen & Land, 1999), whether it is with parents or possibly with
peers and teachers. However, researchers currently define adolescent attachment as an
internal organization, a dramatically different view than attachment as an external
organization as it is during the early years of life. This present view of adolescent
attachment is supported by theory on the purpose of attachment during adulthood. That
is, late in life attachment is primarily utilized as a means of dealing with intense affect,
and therefore should become a characteristic of the individual rather than a construct
dependent upon a relationship with a caregiver (Allen & Land, 1999). Organization of
attachment thus appears to be the most theoretically useful and empirically supported
way in which to define adolescent attachment.
As this view of adolescent attachment organization suggests that attachment is a
critical contributor to internal working models of relationships, it follows that attachment
will have implications for interpersonal interactions in the school environment. When
adolescents exhibit coherence in speaking about and reflecting upon attachment
experiences, these teenagers secure attachment organizations should allow for similarly
coherent views on peer relationships (Allen & Land, 1999). Insecurity should lead to less
coherence in ones thoughts on peer relationships. These deficiencies in the manner in
which peer relationships are perceived will produce poor relationships with others. In the
school environment, a lack of positive peer relationships should drastically impact

Adolescent Attachment 14
academic performance, as children are distracted from their studies by interpersonal
problems.
An increasing body of research suggests that parental relationship factors are
correlated with problems in the school setting. At age four, children who display high
levels of mother-child affect and intimacy are rated by teachers as having greater
academic competency, better work habits, and fewer behavior problems in the classroom.
Furthermore, mother-child dyads exhibiting more positive maternal qualities of
interaction have children with fewer behavior problems and preferred work habits
(Pianta, Nimetz, & Bennett, 1997). The effects that parents have on school performance
may also be evident on a day-to-day basis; although not necessarily indicative of an
insecure attachment relationship, during their late elementary school years, children who
report a high level of parental aversive behavior on one day will have increased peer and
academic problems in their classroom the following day (Repetti, 1996).
There have also been findings to suggest that the security of the attachment
relationship, rather than simply the relationship quality of the parent-child dyad, may
correlate with early school success. In preschool, less secure children display greater
amounts of aggression and anger towards their teachers (DeMulder et al., 2000), which
may impact childrens reception by teachers, motivation to perform well in the classroom,
and consequent scholastic performance. Security is likewise shown to affect teachers
perceptions of children in the classroom, as boys having just completed kindergarten are
reported less likable by their teachers if they are insecurely attached to their parents
(Cohn, 1990).

Adolescent Attachment 15
As children progress through elementary school, the importance of a secure
attachment in contributing to scholastic achievement does not diminish; children at age
seven who have secure parental attachments as assessed by responses to separation
stories and observations have better cognitive performance, a trend which continues
through adolescence (Jacobsen, Edelstein, & Hofmann, 1994). Additionally, secure
attachments during late childhood and early adolescence influence school performance,
as Jacobsen and Hofmann (1997) have found that securely attached children have higher
levels of attention and participation in school, as well as a higher GPAs. Finally,
Finnegan, Hodges, and Perry (1996) theorized that school performance might in turn
influence security, as disruptive antisocial acts performed in the school environment may
cause parental rejection or avoidance in childrens attachment relationships.
Research supports a relationship between attachment and various school
performance outcomes. But are there outside factors that could cause the association
between attachment and school performance? As peer relationships often occur in the
context of the school environment and are often influenced by attachment security, the
quality of ties to ones peers may contribute to the explanation of the relationship
between attachment and academic performance. Hence, the connections between
attachment and peer relations may merit consideration.
Early in development, the nature of the child-caregiver attachment relationship
predicts not how frequently peer interactions are initiated, but rather how peers respond
to the initiator; insecurely attached toddlers begin interactions with peers in ways that
lower the probability of future positive peer interactions (Fagot, 1997). Less secure
preschoolers also display increased aggression and anger towards their peers (DeMulder

Adolescent Attachment 16
et al., 2000), which could account for the impact of security on peer relationships.
Parental relationship qualities, such as mother-child affect and intimacy, are also
correlated with peer social skills in preschoolers (Pianta et al., 1997). Security likewise
affects young childrens models of peer behavior; in an investigation of children in
kindergarten and first grade, Cassidy, Kirsh, Scolton, and Parke (1996) found that
children securely attached to their mothers more often believe that their peers have
positive intent in ambiguous situations. Clearly, a young childs attachment relationship
with his or her caregiver has wide ranging implications for social interactions with peers.
Older childrens attachment also has been found to predict peer relationships.
Secure attachments may foster social competence during middle childhood (Freitag,
Belsky, Grossmann, Grossmann, & Scheuerer-Englisch, 1996), likely impacting a childs
quality and number of friendships; therefore insecurely attached children should have
limited access to positive peer social support. Self-reports of security with ones mother
have been shown to relate to fifth graders acceptance by their peers and their behavior
with friends (Kerns, Cole, & Klepac, 1996), again having implications for the potential
formation of positive peer networks during childhood.
Additionally, the security of a childs parental attachment relationship is related to
qualities within individual peer relationships and not simply the social group as a whole.
Self-reported security and closeness in friendships correlates with perceived parental
attachment (Lieberman, Doyle, & Markiewicz, 1999). Higher peer competence is also
reported for secure adolescents (Allen et al., 1998), and during late adolescence, security
is correlated with peer reports of hostility and self-reports of social support (Kobak &

Adolescent Attachment 17
Sceery, 1988). These studies suggest that peer relationships are affected by the security
of attachment throughout development.
Just as childrens attachment may influence peer relationships, peer relationships
may in turn have drastic influences on school success. Conflict with kindergarten friends
is related to school adjustment problems such as decreased involvement and lower levels
of liking school (Ladd, Kochenderfer, & Coleman, 1996). Also in kindergarten,
increased classroom participation and achievement is correlated with ones number of
friends and peer acceptance (Ladd, Birch, & Buhs, 1999). Peer acceptance similarly
predicts school adjustment in children in the first through third grades (Diehl, Lemerise,
Caverly, Ramsay, & Roberts, 1998).
During adolescence the connection between peer relationships and academic
performance continues. Acceptance by peers in both the sixth and seventh grades is
positively related to the pursuit of prosocial goals and behavior, while negatively related
to the pursuit of academic social responsibility goals (Wentzel, 1994). There are some
more positive influences of peers on scholastic performance, though, as Berndt and Keefe
(1995) have found that children with positive friendship features in the seventh and
eighth grades are more involved in school. Those with negative friendship features are
highest on self-reported disruption, have the lowest school involvement, and seem to be
more affected by these negative features of friendships than their peers are affected by
their own positive friendship features (Berndt & Keefe, 1995). Peer relationships also
appear to affect discipline-specific academic success; middle and high school students
English efficacy is related to high perceived peer attachment as measured by the
Inventory of Parent and Peer Attachment (Laible, Carlo, & Raffaelli, 2000).

Adolescent Attachment 18
Consequently, it may be concluded that school performance during adolescence is
connected to aspects of teenagers peer relationships and social skills.
Although much research has been conducted on the relationship between
attachment and peer relations, as well as each variables connection to school
performance, the relationships among these three factors have not been thoroughly
explored in adolescence. Moreover, the use of a wide range of school adjustment
measures, including self-, peer-, and school-reported variables, to assess academic
performance in the teenage years has not been employed when investigating connections
to both attachment and peer relations. The study of mediators between attachment and
school outcomes is lacking; it has been suggested that this relationship could be mediated
by a childs behavioral self-restraint (Feldman & Wentzel, 1990), but the strong
correlations between attachment and peer relationships naturally lead one to question if
peer relationships are a crucial mediator. Finally, the possibility of peer relationships as a
moderator remains to be examined, as it may be the interaction of attachment with peer
relations that influences school outcomes.
The current study will aim to clarify the relationships between attachment, peer
relations, and school outcomes during adolescence. Specifically, this study will address
the following questions:
1. Are both perceptions of attachment and attachment organization related to
school performance during adolescence?
2. Do perceptions of attachment and attachment organization predict the quality
of adolescents peer relationships?
3. Do peer relationships predict school satisfaction and achievement?

Adolescent Attachment 19
4. Is there any evidence of mediated effects of peer relations?
5. Is there an interaction between perceived attachment to parents and to friends
in predicting school success?
It is hypothesized that attachment perceptions and organization will correlate with
teenagers perceptions of their academic environment, as well as their actual success in
school. Secure attachment organization is also predicted to correlate with the presence of
social support and peer relationships. Positive peer relationship factors are hypothesized
to predict greater satisfaction with and performance in school. Although both attachment
and peer relationships should independently relate to academic variables, with the most
secure adolescents as well as those with reports of social acceptance performing at the
highest level in school, academic outcomes will be explained best by a model taking into
account both attachment and peer relations.
Method
Participants
One hundred and sixty-six ninth and tenth grade students (89 male, 77 female)
participated in this study. Upon entering the study, adolescents ranged in age from 14 to
18.75 years with a mean of 15.9 years (SD = 0.8). The self-identified racial/ethnic
background of the sample was 59.6% European-American, 38.6% African-American, and
1.8% other. Thirty percent of adolescents were living with both biological parents. The
mean family income was $32,030 with a range from less than $2,500 to greater than
$70,000.
Participants were recruited through public school systems serving rural, suburban,
and moderately urban areas. Adolescents were selected for inclusion in the study based

Adolescent Attachment 20
upon the presence of at least one of four possible academic risk factors: 1) failing a single
course for at least a single marking period, 2) any history of grade retention, 3) ten or
more absences in one marking period, or 4) any history of school suspension. These
broad selection criteria were used to include a wide range of adolescents who could be
identified from academic records as having the potential for future academic and social
difficulties, including adolescents already experiencing serious difficulties and those who
were performing adequately with only occasional, minor problems. As intended, these
criteria identified approximately one-half of all ninth and tenth grade students as eligible
for the study.
Each teenager was asked to name several friends who knew him or her well to
participate in a peer interview. Sixty-one male (47.3%) and sixty-eight female (52.7%)
peers were interviewed. The average age of peers participating was 16.4 years (SD = 1.4
years). Seventy-three (56.6%) of the peers were European American, fifty (38.8%) were
African American, and six (4.7%) were of another racial background.
Procedure
After adolescents were identified as meeting the criteria of the study, letters
explaining the study were sent to the families of each potential participant. Interested
families were contacted by phone. If both the teenager and the parent(s) agreed to
participate, they were scheduled for two, three-hour sessions. Approximately 50% of the
approached families agreed to participate. Families were paid a total of $105 for their
participation. Active informed consent was obtained from both adolescents and their
parents.

Adolescent Attachment 21
The teenagers participating in the study were asked to provide names of up to five
friends who knew them well to participate in a peer interview. Up to two of these
friends were contacted for an interview. In cases in which data were gathered from two
peers, their ratings of the teenager in the study were averaged to create one peer variable.
Active consent was obtained from both peers and parents of peers participating in the
study. Peers were paid $10 to come in for a one-hour session, during which they
completed written questionnaires and used Q-sort techniques to rate the teenager who had
nominated them for participation in the study. Data were collected from peers in sessions
separate from those of the participating target teenagers. Transportation and childcare
were provided for all participants as necessary.
Study participants were assured that all information would be kept confidential.
Data were protected by a confidentiality certificate issued by the United States
Department of Health and Human Services, which protected information from subpoena
by federal, state, and local courts.
Measures
Adult Attachment Interview and Q-Set (George, Kaplan, & Main, 1996; Kobak,
Cole, Ferenz-Gillies, Fleming, & Gamble, 1993). Researchers administered this
structured interview to probe individuals descriptions of their childhood relationships
with parents, with requests for abstract terms as well as specific supporting memories.
For instance, participants were asked to list five words describing their early childhood
relationships, and then to describe specific instances that reflected each word. Other
questions addressed specific instances of upset, separation, loss, trauma, and rejection.
Finally, interviewers asked participants to provide more integrative descriptions of the

Adolescent Attachment 22
changes in their relationships with their parents and the current state of these
relationships. Each interview consisted of eighteen questions and lasted on average one
hour. Slight adaptations to the adult version of the attachment interview were made to
make the questions more natural and easily understood by an adolescent population
(Ward & Carlson, 1995). All interviews were audiotaped and then transcribed for coding.
The AAI Q-Set (Kobak et al., 1993). This Q-set was designed to resemble the
Adult Attachment Interview (AAI) Classification System (Main & Goldwyn, in press),
while also yielding continuous measures of qualities of attachment organization. The
data produced by this system nevertheless can be reduced via an algorithm to
classifications that largely agree with three category ratings from the AAI Classification
System (Borman-Spurrell, Allen, Hauser, Carter, & Cole-Detke, 1995; Kobak et al.,
1993). Each coder using the Q-set method reads an AAI transcript and provides a Q-sort
description by using a forced distribution, assigning 100 items into nine categories
ranging from most to least characteristic of the interview. All interviews were blindly
rated by at least two coders with extensive training in both the Q-sort and Main AAI
Classification System.
The Q-sorts were then compared with dimensional prototype sorts for: secure
versus anxious interview strategies, reflecting the overall degree of coherence of
discourse, the integration of episodic and semantic attachment memories, and a clear
objective evaluation of attachment; preoccupied strategies, reflecting either rambling,
extensive, but unfocused discourse about attachment experiences, or angry preoccupation
with attachment figures; dismissing strategies, reflecting inability or unwillingness to
recount attachment experiences, idealization of attachment figures that is discordant with

Adolescent Attachment 23
reported experiences, and lack of evidence for valuing attachment; and deactivating
versus hyperactivating strategies, representing the overall balance of dismissing and
preoccupied styles. Kobak et al. (1993) validated the use of these dimensions, stating
that they accurately capture the constructs of the AAI Classification System.
Each participants scale score consisted of the correlation of the 100 Q-sort items
with each attachment dimension (ranging from 1.00 to 1.00). The Spearman-Brown
reliabilities for the final scale scores were .84 for secure, .89 for dismissing, .82 for
preoccupied, and .91 for hyperactivating versus deactivating scales. Although this system
was designed to yield continuous measures of attachment organizations, rather than
replicate classifications from the Main and Goldwyn (in press) system, the current study
reduced the scale scores to classifications by using the largest Q-scale score above .20 as
the primary classification (Kobak et al., 1993). When scores were compared to a
subsample (N=76) of the adolescent AAIs classified by an independent coder with wellestablished reliability in classifying AAIs (U. Wartner), 74% received identical codes
(kappa = .56, p < .001), and 84% of scores matched in terms of security versus insecurity
(kappa = .68).
Inventory of Parent and Peer Attachment (Armsden & Greenberg, 1987).
Adolescents perceptions of the current degree of trust, communication, and alienation in
their relationships with their mothers and peers were assessed using this 25-item
inventory. Teenagers rated how true each item was with respect to their mother and their
friend(s) on a five-point scale, producing a security composite score. Sample items
include: I trust my mother (trust), My mother encourages me to talk about my
difficulties (communication), and I feel alone or apart when I am with my mother

Adolescent Attachment 24
(alienation). Cronbach alphas measuring internal consistency for the three subscales
were .91, .88, and .86, respectively, and .92 for the composite score. This questionnaire
has been shown to have good test-retest reliability and has been related to other measures
of family environment and teenagers psychological functioning (Armsden & Greenberg,
1987).
Adolescent Self-Perception Profile (Harter, 1988). Participating teenagers and
their peers completed a modified version of the Adolescent Self-Perception Profile
regarding the teenagers social acceptance and close friendship competence. For peer
reports, the same items were used as in the original measure, but were modified for this
peer report measure to allow peer ratings of the adolescent, rather than self-ratings.
Teenagers and peers rated how true each item was of the participating teenagers on a
four-point scale from not true at all to very true. Examples of items from the scales
include: Some people are popular with others their age, but other people are not very
popular (social acceptance) and Some people do not have a really close friend to share
things with, but other people do have a close friend to share things with (close
friendship).
The Adolescent Self-Perception Profile was also used to measure participating
teenagers scholastic competence. Only peers reports were used for this scale. As with
the social acceptance and close friendship scales, peers were asked to rate how true the
items were of their friend. An example of an item from this scale is Some people do
very well at their class work.
For the peer reports of teenagers with more than one friend participating as their
peer in the study, scores from the two peers who rated the teenager were averaged to

Adolescent Attachment 25
create a single rating for each scale. The untrained adolescent raters produced ratings that
correlated with each other surprisingly well (Spearman-Brown r = 0.63), and the resulting
scale had good internal consistency (Cronbachs = 0.84).
Annual Weighted GPA. Scholastic records were obtained from the participating
teenagers schools. Each students grade point average for the full year was computed,
adjusting for any applied/special or advanced/honors courses the participants were taking.
Advanced/honors courses were awarded an extra point in the GPA computation, whereas
applied/special courses were given a half of a point deduction.
Grades were also adjusted so that the number of credits awarded by the school for
each course (.5, 1) was taken into consideration. All classes, not simply the core
classes of English, science, social studies, and math, were included in the GPA
computation.
School Risk. A summary school risk variable was created to account for
disciplinary action taken against participating teenagers. To form the variable, the
number of times the teenager had been suspended, the number of times the teenager had
been expelled, and whether or not the teenager had dropped out of school were taken into
account. This information was obtained directly from the teenagers schools. The
categorical school risk variable ranged from zero (low school risk) to five (high school
risk). For a teenager to receive a score of zero on the school risk variable, he or she must
have never been suspended. School risk values of one to four were received by teenagers
with varying numbers of suspensions, with level one teenagers being suspended one or
two times, level two teenagers suspended three to five times, level three teenagers
suspended six to ten times, and level four teenagers suspended eleven or more times. A

Adolescent Attachment 26
level five school risk was assigned to teenagers who had been expelled from school at
any point during their education or had dropped out of school.
Total Attachment to School. A measure of attitudes towards school was adopted
from the Quality of School Life Scale (QSL) by Epstein and McPartland (1976/1978) and
administered to the participating teenagers. The questionnaire was composed of twentysix questions related to the adolescents experiences at his or her current or most recent
school. Questions one to fourteen were true/false questions, questions fifteen to twentytwo were multiple choice questions, and numbers twenty-three to twenty-six included
statements rated on a Likert scale. The multiple choice and Likert scale questions were
rescaled to dichotomous scores.
Each question loaded on one of three subscales based on the three dimensions of
the quality of school life. The satisfaction with school scale examined students general
reactions to school, the commitment to class work scale measured students level of
interest in their schoolwork, and the reactions to teachers scale had students evaluate their
academic and personal interactions with teachers. Each subscales score was the mean of
the subscale items times the number of items in the scale. No more than one of the five
items of the satisfaction with school scale could be missing in order to compute a
participants score on this scale. Up to three of the eleven items on the commitment to
class work scale and up to three of the ten items on the reactions to teachers scale could
be missing to compute an individuals scores on these scales. To obtain one measure of
attachment to school, all three scales were combined by taking the mean of the responses
to all twenty-six questions and multiplying by twenty-six. No more than four items could
be missing to compute this rating of total attachment to school.

Adolescent Attachment 27
Results
Preliminary Analyses
Demographic factors. Means and standard deviations of all demographic factors
are presented in Table 1. Correlations were run between predictor variables and the
outcome variables for the full sample (Tables 2, 3, and 4).

Table 1
Demographic variables for target teenagers and peers

Age (in years)


Mean
(SD)
Gender
Male
Female
Race/Ethnicity
European American
African American
Other

Table 2

Teenagers

Peers

15.9
.8

16.4
1.4

53.6%
46.4%

47.3%
52.7%

59.6%
38.6%
1.8%

56.6%
38.8%
4.7%

Adolescent Attachment 28
Correlational analyses of attachment and school factors
I. Attachment organization as measured by the Adult Attachment Interview correlated
with school factors
Annual
Weighted GPA

Scholastic
Competence

Attachment
Secure
.23**
.10
Preoccupied
-.07
-.19*
Note: ***p.001, **p.01, *p.05, +p.10.

School Risk

Attachment to
School

-.25**
.08

.12
-.00

II. Perception of attachment to ones mother as measured by the Inventory of Parent and
Peer Attachment correlated with school factors
Annual
Weighted GPA

Scholastic
Competence

Attachment
Maternal
-.04
.06
Note: ***p.001, **p.01, *p.05, +p.10.

Table 3

School Risk

Attachment to
School

-.11

.19*

Adolescent Attachment 29
Correlational analyses of attachment and peer relationship factors
I. Attachment organization as measured by the Adult Attachment Interview correlated
with peer relationship factors
Attachment
to Friends

Social
Acceptance

Attachment
Secure
.21+
.14+
Preoccupied
-.23+
-.11
Note: ***p.001, **p.01, *p.05, +p.10.

Social
Acceptance
(peer report)

Close
Friendship

Close
Friendship
(peer report)

.18*
-.08

.21**
-.16*

.14
.03

II. Perception of attachment to ones mother as measured by the Inventory of Parent and
Peer Attachment correlated with peer relationship factors
Attachment
to Friends

Social
Acceptance

Attachment
Maternal
.27*
.22**
Note: ***p.001, **p.01, *p.05, +p.10.

Table 4

Social
Acceptance
(peer report)

Close
Friendship

Close
Friendship
(peer report)

.14+

.27***

.11

Adolescent Attachment 30
Correlational analyses of peer relationship and school factors
Annual
Weighted GPA

Scholastic
Competence

Attachment to
.08
.05
Friends
Social
-.09
-.03
Acceptance
Social
Acceptance
.01
.16+
(peer report)
Close
-.01
-.02
Friendship
Close
Friendship
.22*
.32***
(peer report)
Note: ***p.001, **p.01, *p.05, +p.10.

School Risk

Attachment to
School

-.32**

.39**

-.06

-.09

.08

-.03

-.20**

.07

-.16+

.11

Primary Analyses
Analytic strategy. Previous research has linked family relationship characteristics
to quality of peer relationships and school performance; likewise, peer relationship
characteristics have been correlated with academic achievement. Attachment variables
were therefore included as predictors in two of the regression analyses, whereas peer
relationships were used as the predictor for one analysis. Hierarchical regression models
were used for all analyses, with the demographic variables entered into the model first.
The primary variables of interest for the first set of analyses, adolescents perceptions of
attachment and attachment organization, were entered next to examine the variance in
school factors beyond that explained by the three demographic variables (gender, race,
and total family income).
Similarly, for the analyses involving attachment and peer factors, attachment
measures were used as predictors in hierarchical regression models. Finally, hierarchical

Adolescent Attachment 31
regression models were run with peer factors used to examine the variance in academic
performance beyond that explained by the demographic variables. All regressions were
run to control for participants gender, race, and total family income; very few trends and
significant correlations became nonsignificant after accounting for these factors. Results
from these regression analyses are presented with the main result data.
As mediator effects were also hypothesized, separate regression analyses were run
to investigate the role of peer relationship factors as mediators of the relationship
between attachment and school variables.
In the final set of analyses, interaction terms of attachment with peer variables
were entered into hierarchical regression models to determine their contribution in
predicting school success.
Adolescent attachment and school factors. In the first set of analyses,
adolescents perceptions of attachment and organization of attachment were investigated
as predictors of academic variables. Attachment was regressed on the school factors.
Beyond the effects of the demographic variables, a trend towards a main effect was found
for adolescents secure attachment organization and school risk ( = -.15, p .10),
implying that a secure attachment somewhat accounts for lowered school risk. A
significant main effect was found for preoccupied attachment organization and peerreported scholastic competence ( = -.25, p .01), such that lowered scholastic
competence was predicted by preoccupied attachment. Finally, perceptions of attachment
to the adolescents mother were related to school risk ( = -.17, p .05) and attachment to
school ( = .22, p .05). Therefore, those adolescents who perceived greater attachment
to their mothers had stronger attachments to school and a lessened likelihood of being

Adolescent Attachment 32
suspended, expelled, or dropping out of school. Regression analyses of the attachment
and school factors are presented in Tables 5 and 6.

Adolescent Attachment 33

Table 5
Hierarchical regression analyses of attachment organization as measured by the Adult Attachment Interview predicting academic
functioning after accounting for gender, race, and total family income
Annual Weighted GPA
Predictors
I:
Gender
Race
Income

Scholastic Competence

School Risk

Attachment to School

.01
.25**
.34***

.27***

.27***

.08
.01
-.11

.02

.02

-.27***
-.05
-.21*

.12***

.12***

.01
.09
-.07

.01

.01

.01

.13

.02

.01

.01

-.03

.01

II:
Security
.07
.27***
0
.15
.03
.01
-.15+
.03
III:
Preoccupation
-.10
.28***
.01
-.25**
.08*
.06**
.11
.13***
Note: ***p.001, **p.01, *p.05, +p.10. weights are from variables first entry into model.

Adolescent Attachment 34

Table 6
Hierarchical regression analyses of perception of attachment to ones mother as measured by the Inventory of Parent and Peer
Attachment predicting academic functioning after accounting for gender, race, and total family income
Annual Weighted GPA
Predictors
I:
Gender
Race
Income

Scholastic Competence

School Risk

Attachment to School

.03
.25**
.33***

.25***

.25***

.09
.04
-.10

.02

.02

-.27***
-.01
-.22**

.11***

.11***

.02
.10
-.06

.01

.01

.03*

.22**

.05

.04**

II:
Perception of
Maternal
.06
.26***
.01
.07
.02
0
-.17*
.14***
Attachment
Note: ***p.001, **p.01, *p.05, +p.10. weights are from variables first entry into model.

Adolescent Attachment 35
Adolescents attachment and relations to peers. In the next set of analyses,
adolescents attachment perceptions and organizations were examined as predictors of
peer relationship quality. Adolescents peer relationships (attachment to friends, social
acceptance, and close friendships) were regressed on adolescent attachment in separate
regression equations. After accounting for the effects of the demographic variables, main
effects were found for secure attachment organization as a predictor of self-reported
social acceptance ( = .18, p .05), peer-reported social acceptance ( = .29, p .01),
and self-reported close friendship ( = .18, p .05). This indicates that security of
attachment organization was associated with greater social acceptance as rated by the
teenager and his or her peers, as well as the teenagers own perception of having a close
friendship. An additional main effect was found for preoccupied attachment organization
and self-reports of a close friendship ( = -.19, p .05), such that the presence of a close
friendship was less likely to be reported by adolescents with preoccupied organizations.
Lastly, perceptions of attachment to ones mother were related to peer relationship
qualities, as main effects were found for perceptions of attachment to ones mother and
perceptions of attachment to friends ( = .30, p .05), self-reported social acceptance (
= .21, p .05), and self-reported close friendship ( = .29, p .001). In addition, trends
were found for the perception of secure attachment to ones mother predicting peerreported social acceptance ( = .15, p .10) and peer-reported close friendship ( = .16, p
.10). Therefore, like the results for attachment organization, perceptions of attachment
predicted a wide range of peer relationship factors (both self- and peer-reported). Results
for the attachment and peer relationship analyses are presented in Tables 7 and 8.

Adolescent Attachment 36

Table 7
Hierarchical regression analyses of attachment organization as measured by the Adult Attachment Interview predicting peer
relationship factors after accounting for gender, race, and total family income
Attachment to Friends

Predictors
I:
Gender
Race
Income

Social Acceptance

Social
Acceptance
(peer report)

Close Friendship

Close Friendship
(peer report)

.21+
-.05
.12

.05

.05

.06
-.07
-.05

.02

.02

.12
-.07
-.18

.06+

.06+

.15+
.03
.03

.03

.03

.24**
.07
-.08

.07*

.07*

Security
.15
.07
.02
.18*
.04
.02*
.29**
.13**
.07**
III:
Preoccupation
.24+
.10
.05+
-.11
.03
.01
-.09
.07+
.01
Note: ***p.001, **p.01, *p.05, +p.10. weights are from variables first entry into model.

.18*

.06+

.03*

.13

.08*

.01

-.19*

.06+

.03*

-.03

.07+

II:

Adolescent Attachment 37

Table 8
Hierarchical regression analyses of perception of attachment to ones mother as measured by the Inventory of Parent and Peer
Attachment predicting peer relationship factors
Attachment to Friends

Predictors
I:
Gender
Race
Income

Social Acceptance

Social
Acceptance
(peer report)

Close Friendship

Close Friendship
(peer report)

.23*
-.04
.16

.07

.07

.03
-.05
-.06

.01

.01

.10
-.04
-.17+

.04

.04

.15+
.01
.02

.02

.02

.26**
.09
-.10

.08**

.08**

.30**

.15**

.08**

.21**

.05+

.04**

.15+

.07+

.03+

.29***

.10***

.08***

.16+

.11**

.03+

II:
Perception of
Maternal
Attachment

Note: ***p.001, **p.01, *p.05, +p.10. weights are from variables first entry into model.

Adolescent Attachment 38

Adolescent Attachment 39
Adolescents peer relationships and academic factors. Adolescents peer
relationships were examined as predictors of school factors in the third set of analyses.
The academic factors investigated were GPA, peer-reported scholastic competence,
school risk, and attachment to school. School factors were regressed on peer
relationships. After accounting for demographic variables, several main effects reached
significance. These main effects involved attachment to friends and school risk ( = -.26,
p .05), attachment to friends and scholastic competence ( = .09, p .05), attachment to
friends and total attachment to school ( = .44, p .001), self-reported close friendship
and school risk ( = -.16, p .05), peer-reported close friendship and GPA ( = .18, p .
05), and peer-reported close friendship and peer-reported scholastic competence ( = .32,
p .001). Additionally, a main effect for peer-reported social acceptance and peerreported scholastic competence ( = .21, p .05) approached significance.
These data indicate that better relationships with peers predict various school
outcomes. No significant main effects were found for social acceptance and the
academic variables, although trends were noted. This result suggests that school
outcomes are not as closely related to social acceptance as attachment to friends and both
self- and peer-reported close friendships. Table 9 presents the regression analyses for the
peer relationship and school factors.

Adolescent Attachment 40

Table 9
Hierarchical regressions predicting school factors from peer factors after accounting for gender, race, and total family income
Annual Weighted GPA
Predictors
I:
Gender
Race
Income

Scholastic Competence

School Risk

Attachment to School

.02
.16
.33*

.20**

.20**

.11
-.01
-.11

.02

.02

-.17+
.05
-.34**

.12**

.12**

.03
.06
-.25+

.05

.05

.06

.20**

.09*

.03

.01*

-.26*

.18**

.06*

.44***

.24**

.19***

.03
.26**
.32***

.26***

.26***

.10
.03
-.11

.02

.02

-.28***
-.01
-.21*

.11***

.11***

.02
.10
-.08

.01

.01

-.06

.27***

.01

-.04

.02

-.07

.12***

.01

-.09

.02

.01

-.04

.26***

-.03

.02

-16*

.14***

.03*

.06

.01

II:
Attachment to
Friends
I:
Gender
Race
Income
II:
Social Acceptance
III:
Close Friendship

Table continued on next page

Adolescent Attachment 41

Table 9 (continued)
Annual Weighted GPA
Predictors
I:
Gender
Race
Income

Scholastic Competence

School Risk

Attachment to School

.03
.24*
.27**

.20***

.20***

.10
.03
-.11

.02

.02

-.32***
-.08
-.15

.15***

.15***

.04
.12
-.13

.02

.02

-.02

.02

.10

.03

.01

II:
Social Acceptance
.06
.20***
0
.16+
.05
.03+
.06
.15***
(peer report)
III:
Close Friendship
.17*
.23***
.03*
.32***
.11**
.09***
-.09
.15***
(peer report)
Note: ***p.001, **p.01, *p.05, +p.10. weights are from variables first entry into model.

Adolescent Attachment 42
Peer relationships as a mediator between attachment and school factors. In a
series of analyses, attachment to peers and the presence of close friendships were
examined as mediating the connection between attachment organization and school
variables (both school risk and peer-reported scholastic competence). According to
Baron and Kenny (1986), a mediator must meet four conditions in four respective
equations: 1) the mediator must be affected by the independent variable, 2) the dependent
variable must be affected by the independent variable, 3) the dependent variable must be
affected by the mediator, and 4) the effect of the independent variable on the dependent
variable must be greater than the effect of the mediator on the dependent variable.
Possible mediation pathways explaining the relationship among attachment, peer
relationships, and academic achievement were deduced from the analysis of both
uncontrolled and controlled correlations. Three pathways involving the uncontrolled
variables (secure attachment organization perceived peer attachment school risk;
secure attachment organization close friendship school risk; preoccupied
attachment organization close friendship scholastic competence) were identified for
further analysis, as was one pathway involving the controlled variables (secure
attachment organization close friendship school risk).
Peer attachment perceptions were found to mediate the relationship between
secure attachment organization and school risk only when regression equations did not
control for race, gender, and total family income. When demographic controls were not
considered, the connection between secure attachment organization and school risk was
weakly mediated by self-reported close friendship. The other two mediator analyses did
not yield significant results. These results are presented in Table 10.

Adolescent Attachment 43
Table 10
Analyses of peer relationship variables as mediators of the connection between
attachment and school variables using Baron and Kenny tests
Attachment Attachment Peers and Change in
and School and Peers
School
Regression?

Evidence of
a Mediated
Effect?

Attachment Measure:
Secure AAI
School Measure:
Not
-.25**
.21+
-.32**
Yes
School Risk
significant
Peer Measure:
Peer Attachment
Attachment Measure:
Secure AAI
School Measure:
Yes,
School Risk
-.25**
.21**
-.20**
-.22**
although
Peer Measure:
weak
Close Friendship
(peer report)
Attachment Measure:
Preoccupied AAI
School Measure:
Scholastic
Competence
-.19*
-.16*
.32***
-.20*
No
(peer report)
Peer Measure:
Close Friendship
(self-report)
(Controlled)
Attachment Measure:
Secure AAI
School Measure:
Not
-.15+
.18*
-.16*
No
School Risk
significant
Peer Measure:
Close Friendship
(self-report)
Note: ***p.001, **p.01, *p.05, +p.10. weights are from variables first entry into
model.

Adolescent Attachment 44
Interactions of attachment and peer relationships. Adolescents perceptions of
attachment to mothers and to peers were investigated to examine whether they might
interact to predict total attachment to school. A significant interaction was found ( = .
25, p .05) and is depicted in Figure 1. The lowest attachment to school was found for
teenagers who had perceptions of high attachment to mothers and low attachment to
peers. Slightly greater, yet still low, attachment to school was found for adolescents with
perceptions of both low attachment to mothers and to peers. This group was relatively
close on level of attachment to school to those teenagers with low maternal attachment
and high peer attachment. Comparatively high levels of attachment to school were found
for adolescents with high perceptions of attachment to mothers and to peers.

Adolescent Attachment 45
Figure 1
Interaction between perception of attachment to ones mother and perception of
attachment to peers resulting in varying levels of total attachment to school

Attachment to School

Perception of High
Maternal Attachment

0.5
0
Perception of Low
Maternal Attachment

-0.5
-1
-2

-1

Perception of Attachment to Peers

Overall, the data collected in this study suggest three predictive relationships: 1)
attachment predicting school performance, 2) attachment predicting peer relationships,
and 3) peer relationships predicting school performance. Both perceptions of maternal
attachment and attachment organization predicted academic factors. Likewise, regression
equations support that both perceptions of secure maternal attachment and secure
attachment organization predict greater peer relationships. Additionally, greater peer
relationships predicted more positive school outcomes. Some evidence of perceived peer
attachment serving as a mediator between secure attachment organization and school risk
was found, but only when demographic influences were not considered in the models.
Also supported was a significant interaction between perceptions of attachment to ones
mother and to peers resulting in varying levels of attachment to school.

Adolescent Attachment 46
Discussion
This study provides evidence for the independent links between adolescent
attachment, peer relationships, and school factors, while also providing support for the
possibility of peer relationships serving as a mediator between attachment and school
variables. An interaction among these three domains is also supported, with varying
levels of perceived attachment to ones mother and perceived attachment to peers
predicting attachment to school.
Connections between attachment and school performance were hypothesized, and
data from this study suggest that such relationships exist. Attachment organization
predicted several different aspects of school success, as adolescents with secure
attachment organizations were less likely to encounter disciplinary difficulties. Also,
teenagers with preoccupied attachment organizations had lower peer-reported scholastic
competence. Additionally, ones perception of maternal attachment predicted less
likelihood of disciplinary action.
Not all school variables could be predicted from individuals attachment,
however, with both attachment to school and GPA not being predicted by maternal
attachment perceptions or attachment organization. This result could be accounted for if
one assumes that teenagers specific relationships within the academic environment (such
as those with only their school peers and teachers) most greatly affect satisfaction in the
school environment. This conclusion follows from the assumption that perceptions of
attachment with parents and peers and attachment organization do not necessarily
represent teenagers general relationship quality. However, attachment theory predicts

Adolescent Attachment 47
that attachment organization should affect ones internal model of all relationships, so this
finding remains somewhat surprising.
The lack of support for GPA being predicted by attachment variables can be
explained, though, by this studys omission of an analysis of certain factors intuitively
related to scholastic performance. For example, teenagers IQ, the quality of the school
and home environments, and additional family factors such as parent emphasis on
education were not investigated. These factors relationship to academic achievement
should be considered in future studies. Additional work could also study attachment and
school performance while controlling for these additional factors, perhaps allowing a
closer relationship between attachment and school performance to be found.
In the analyses of attachment organization and peer factors, secure and
preoccupied attachments predicted several peer relationship variables. As hypothesized,
adolescents with more secure attachment organizations were perceived by themselves and
their friends to have high levels of peer group acceptance. This finding is congruent with
attachment theory, as those individuals with secure attachment organizations should have
an internal working model that allows them to foster the development of positive
relationships with others. Teenagers with preoccupied attachment organizations were less
likely to report having a close friendship, also as predicted by attachment theory, as an
insecure internal working model should make close interpersonal relationships both
harder to develop and more difficult to perceive in ones environment (Allen et al., 1998;
Bowlby, 1973/1977/1989; Cassidy, Kirsh, Scolton, & Parke, 1996; Dodge, 1993; Slough
& Greenberg, 1990). However, preoccupied attachment organization was not found to
predict teenagers social acceptance, or peers perceptions of preoccupied teenagers close

Adolescent Attachment 48
friendships. Therefore, a preoccupied attachment organization, which is indicative of an
insecure internal working model, was most closely associated with self-perceptions of
close friendships. As attachment organization should be most closely associated with
attachment-like, or close, relationships, relationships with close friends rather than the
peer group as a whole would be expected to be more directly related to insecure working
models. This studys findings support this assertion.
Perceptions of attachment were likewise predictive of teenagers quality of peer
relationships. Attachment perceptions were most strongly linked with self-reported social
acceptance and close friendships. It is possible that individuals tendencies to view the
world in either a positive or negative light contributed to these results. Having found that
teenagers self-reports of attachment and self-reports of peer measures correlate may be
due to reporter bias effects. However, trends were noted for perceptions of attachment
predicting peer-reported social acceptance and the presence of a close friendship. These
trends suggest that self-perceptions of attachment may be related to teenagers peer
relationships above and beyond the effect due to all measures coming from the same
reporter. Perhaps if future studies utilize other measures of social acceptance and close
friendship, such as observational reports, the discrepancy between self- and peerreported data would be resolved, and the linkage between perceptions of attachment and
peer relationship quality could be clarified.
Self- and peer-reported peer relationship quality were also found to relate to
academic factors, predicting a wide range of school outcomes. Teenagers self-reported
attachment to their friends predicted how frequently teenagers had discipline problems as
well as adolescents attachment to school. Low levels of discipline problems were

Adolescent Attachment 49
additionally predicted by teenagers reports of close friendships. These findings suggest
that having a close friend may serve as a protective factor against school discipline
problems, as those teenagers who believe they have a reliable friend may be less likely to
act out in the classroom or submit to negative peer pressure, actions that may result in
school suspensions or expulsions. Also, peer-reported close friendships were predictive
of peer-reported scholastic competence as well as students GPA, as was hypothesized.
This finding supports the assertion that students doing well in school are more likely to
have a close friend.
As only trends were noted for social acceptance variables, acceptance by ones
peer group was not an important peer predictive factor for scholastic success. Perhaps
this is because adolescents may need only one good friendship to protect them from low
scholastic performance and dissatisfaction with school, rather than wide peer acceptance.
The results could also be accounted for by the varying influences of peer groups; both
positive and negative peer pressure may influence school performance, with the negative
effects of being close to ones peer group offsetting the positive effects for some
teenagers.
Regression analyses suggest that peer relationship factors may serve as mediators
between attachment and school variables in a couple specific instances. It is only when
attachment organization is used that mediated effects were found, suggesting that security
of attachment organization, not perceptions of attachment, is crucial in the mediator
relationship. Furthermore, significant mediated effects only occur when gender, race, and
total family income are not entered into the regression equations, as these demographic
variables are significantly confounded with the mediated relationships. Under these

Adolescent Attachment 50
conditions, the connection between secure attachment and lower levels of discipline
problems is mediated by perceptions of peer attachment. That is, the relationship
between security and school outcomes becomes nonsignificant once the effect of
teenagers attachment to their peers is considered. The relationship between security and
disciplinary action is also weakly mediated by peer reports of close friendships. Hence,
although only two mediated effects were found, data do suggest that in some specific
situations, peer relationship quality may act as a mediator for the relationship between
attachment and school outcomes.
An alternative explanation for the presence of only two mediated effects is that
perhaps, in general, peer relationships do not serve as a mediator for the connection
between attachment and school factors. It is possible that one piece of the attachment
construct predicts peer relationship quality, another piece of the attachment construct
predicts school outcomes, and that these two pieces of attachment are non-overlapping
aspects of the construct. This would account for the lack of mediated effects for all of the
controlled relationships investigated and would explain the independent connections
between the three constructs (attachment, peer relationships, and school outcomes) being
investigated.
One final relationship between the three constructs was found; an interaction was
supported with perceived attachment to mothers and perceived attachment to peers
predicting students satisfaction with school, commitment to class work, and reactions to
their teachers. High perceived attachment to mothers and low perceived attachment to
peers predicted the lowest attachment to school. This result suggests that adolescents can
be divided into distinct groups based on perceived maternal and peer attachment, and that

Adolescent Attachment 51
these groups have varying levels of school attachment. There is a group of teenagers who
are happiest at home, feeling a strong connection to their mothers and yet, because of not
feeling connected to their peers, are struggling in the school environment. That is, the
data suggest that adolescents with high levels of perceived attachment to parents and
peers are likely to be satisfied with school, except in cases where high levels of perceived
attachment to ones mother are not balanced by high levels of perceived attachment to
peers. For most adolescents, though, the greater the perceived attachment to ones
mother and peers, the higher the level of attachment to school, as was hypothesized. For
all teenagers, higher perceived attachment to peers predicted a greater connection to
school; it seems that the effect of high perceived attachment varies only for teenagers
perceived maternal attachment.
One interesting conclusion from these mediator and moderator results has to do
with the differences in the findings investigating attachment perceptions and attachment
organization. Although the links between attachment and peers as well as those between
attachment and school were evident when using both the IPPA and AAI, this was not the
case for the mediated and moderated relationships. Findings of this study support peer
relationships serving as a mediator between attachment organization and school
measures. No such relationships were found for perceptions of attachment. This may be
a result of attachment organization unconsciously affecting peer relationships, which in
turn may predict school success. Whatever its cause, simply perceiving that ones
relationship with attachment figures is secure is not sufficient to foster peer relationships
and their resulting positive effects on academic life. Similarly, a perception of a negative

Adolescent Attachment 52
parental relationship may not be damaging to peer relations and school life if one can
reflect on these experiences coherently, hence forming a secure attachment organization.
Although attachment organization may be crucial in discovering mediated
relationships, it is also plausible that the lack of mediators found in the investigation of
attachment perceptions could be indicative of a lack of true peer mediators. It is possible
that some other variable is mediating the connection between home life and scholastic
performance, and that the few cases in which mediated relationships were found were
simply due to this other association. Hence, the need for further work in this area is
evident.
Even though attachment organization could be used as a predictor variable in
significant mediated relationships, it was only perceptions of attachment that were found
to interact with peer factors to produce varying levels of attachment to school. This is
evidence that it is teenagers view of their relationships, rather than actual attachment
organization, that is necessary in the moderated relationship among parent-child relations,
peer relations, and school attachment. That is, the teenagers who feel the lowest
attachment to school may only perceive positive relationships with their mothers and
poor relationships with their peers; however, these perceptions do affect their actual
satisfaction with school.
There are several implications of the findings of this study. The lack of strong
evidence for the mediating or moderating role of all of the peer relationship factors
investigated suggests that there could be other facets of the peer relationship not
investigated in this study that may explain connections between attachment and school
factors. It is also possible that some factor that is not related to adolescents peer

Adolescent Attachment 53
relations could serve as a primary mediator or moderator between attachment and school
outcomes. For example, teacher-child relationships, parenting style, and quality of the
home environment could be studied in the future, clarifying these factors roles in linking
parent-child relationships to adolescents school satisfaction and performance.
Also, this study has a few inconsistencies in its findings, particularly
discrepancies among the data from different reporters and among the different measures
of peer and school constructs (that is, peer attachment, social acceptance, and the
presence of a close friendship; and GPA, scholastic competence, school risk, and
attachment to school). The inconsistencies across reporters suggests the need for
observational peer and school data, or teacher reports rather than self- and peer-reported
ratings. Furthermore, future directions of research may investigate why certain factors
within the peer and school constructs were better predictor, mediator, and moderator
variables.
There are also some limitations of this studys design that require attention. First,
the ability to generalize findings from the sample is in question. A group of academically
at-risk teenagers were recruited as study participants. It is unclear whether the links
between the three constructs, as well as the mediated and moderated effects found in this
study, would hold with a more inclusive adolescent population. However, the use of the
limited sample was beneficial in some respects, as it allowed for data collection from a
sample of individuals with a wide range of scholastic difficulties, helping to provide
meaningful results. Additionally, there was diversity in the perceptions of attachment and
attachment organizations of the participating teenagers which also allowed for sufficient
variance within the sample.

Adolescent Attachment 54
An additional limitation of this study is its difficulty in determining theoretically
meaningful definitions of peer relationship factors and adolescent attachment. Close peer
relationships may mean different things to different teenagers, and the actual structure of
adolescents friendships may vary greatly. Furthermore, the impact of social groups and
individual friendships on development varies among teenagers, with some peers
reinforcing academically prosocial goals, while others undermining school success.
Future work may wish to address these issues, as well as examine how in some cases
peers may serve as attachment figures for one another.
In addition to the limitations of the peer relationship factors, problems also
emerge with the construct of attachment. Adolescent attachment is an emerging area of
study. Researchers have presented differing ways in which adolescent attachment may be
conceptualized; for example, attachment may be defined as perceptions of attachment
within adolescent relationships (Bowlby, 1989) or attachment organization (Main et al.,
1985). The most optimal manner in which to measure an individuals attachment security
also remained to be determined. Hence, the validity of present measures of attachment is
questionable. Additionally, Kobak and Sceery (1988) have suggested that the Adult
Attachment Interview (AAI) may be lacking in validity except when used with
individuals in late adolescence or adulthood.
Unfortunately, the validity of the other measure of attachment used in this study,
the Inventory of Parent and Peer Attachment (IPPA), is also questionable if used as a
measure of attachment organization (rather than just as a measure of perceptions of the
parent-child relationship). As the IPPA measures perceived closeness, a report of a
negative relationship with ones parents on the IPPA is not necessarily indicative of an

Adolescent Attachment 55
insecure relationship. In fact, being able to reflect on negative life experiences in an
organized fashion, allowing for the acknowledgement of the impact of disturbing
episodic memories, and yet integrating these memories into a coherent whole, is
characteristic of security. An individual who does reflect in this manner would be
categorized as having a secure attachment organization by the AAI, and yet would be low
on security according to the IPPA. Likewise, denial of the importance of negative events
in ones past, as evidenced by the presence of negative episodic memories but positive
descriptions of ones childhood, would result in a secure classification on the IPPA, but is
evidence of an insecure attachment organization. Thus, the IPPA is not necessarily a
valid measure of attachment, but rather an indicator of various perceived characteristics
and qualities of the parent-child relationship. This is the capacity in which the IPPA was
used in this study and it is therefore crucial to recognize the differences in results from
the IPPA versus the AAI. When investigating attachment organization and perceptions of
attachment, the differences in the IPPA and AAI make the results from these two
attachment measures difficult to combine into one attachment construct. Future study
could focus on the area of adolescent attachment measurement, hopefully eliminating
some of the difficulties in researching this construct.
Lastly, there are important limitations in the way that the construct of positive
school outcomes was defined. There are several different ways that teenagers, parents,
and teachers may conceive adolescent school success, such as happiness in the school
environment, classroom participation, and the obtainment of advanced degrees. In this
study, success in school was defined in terms of teenagers grades, peer-reported success
in school, disciplinary actions, and self-reports of connections to school. Despite the care

Adolescent Attachment 56
given to the selection of these school variables, it is possible that these measures did not
encompass school success in the most meaningful way. If this is the case, perhaps future
work could investigate the additional ways in which school success can be defined, so as
to see if in clarifying the definition stronger results are obtained.
In conclusion, despite its limitations, the present study provides support for the
independent links between attachment perceptions and organization, peer relationships,
and school factors during adolescence. Also supported is the possible role of peer
relationships as a mediator (when investigating attachment organization) or moderator
(when investigating perceptions of maternal attachment). Additional research is
suggested in this area, however, so as to clarify the impact of additional family and
school environment factors that may influence academic achievement. More precise
conceptualizations of the three constructs would also help further this area of study,
allowing for greater precision in defining exactly what elements of the parent-child and
peer relationships predict meaningful aspects of school success.
A better understanding of the relationships among attachment, peer relations, and
school outcomes is valuable as it may serve to focus interventions for academically atrisk adolescents. As academic success does influence a wide range of developmental
outcomes, possible findings could positively affect many aspects of teenagers lives.
With the recognition that attachment perceptions, attachment organization, and peer
relations affect educational outcomes, the importance of developing academic
interventions that extend beyond classroom scholastic instruction becomes evident.
Hopefully the relationships among the constructs of attachment, peer relations, and
school success will soon be more thoroughly understood. Such knowledge could allow

Adolescent Attachment 57
for the development of successful intervention programs which, through targeting aspects
of attachment and peer relationships, may result in higher levels of educational and
developmental success for teenagers.

Adolescent Attachment 58
References
Allen, J. P., Moore, C., Kuperminc, G., & Bell, K. (1998). Attachment and
adolescent psychosocial functioning. Child Development, 69(5), 1406-1419.
Allen, J. P., & Land, D. (1999). Attachment in adolescence. [Chapter] Cassidy, J.
& Shaver, P. R. (Eds.). Handbook of attachment: Theory, research, and clinical
applications. (pp. 319-335). New York: The Guilford Press. xvii, 925 pp.
Armsden, G. C., & Greenberg, M. T. (1987). The inventory of parent and peer
attachment: Individual differences and their relationship to psychological well-being in
adolescence. Journal of Youth and Adolescence, 16(5), 427-454.
Berndt, T. J., & Keefe, K. (1995). Friends influence on adolescents adjustment
to school. Child Development, 66(5), 1312-1329.
Borman-Spurrell, E., Allen, J. P., Hauser, S. T., Carter, A., & Cole-Detke, H. C.
(1995). Assessing adult attachment: A comparison of interview-based and self-report
methods. Manuscript submitted for publication.
Bowlby, J. (1969). Attachment and loss, Vol. 1: Attachment. New York: Basic
Books.
Bowlby, J. (1973). Attachment and loss, Vol. 2: Separation. New York: Basic
Books.
Bowlby, J. (1977). The making and breaking of affectional bonds: I. Aetiology
and psychopathology in the light of attachment theory. British Journal of Psychiatry, 130,
201-210.
Bowlby, J. (1980). Attachment and loss, Vol. 3: Loss. New York: Basic Books.

Adolescent Attachment 59
Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human
development. New York, Basic Books.
Bowlby, J. (1989). The making & breaking of affectional bonds. New York:
Routledge.
Cassidy, J., Kirsh, S. J., Scolton, K. L., & Parke, R. D. (1996). Attachment
representations of peer relationships. Developmental Psychology, 32(5), 892-904.
Chassin, L., Presson, C. C., Sherman, S. J., Montello, D., & McGrew, J. (1986).
Changes in peer and parent influence during adolescence: Longitudinal versus crosssectional perspectives on smoking initiation. Developmental Psychology, 22(3), 327334.
Cohn, D. A. (1990). Child-mother attachment of six-year-olds and social
competence at school. Child Development, 61, 152-162.
DeMulder, E. K., Denham, S., Schmidt, M., & Mitchell, J. (2000). Q-sort
assessment of attachment security during the preschool years: Links from home to
school. Developmental Psychology, 36(2), 274-282.
Diehl, D. S., Lemerise, E. A., Caverly, S. L., Ramsay, S., & Roberts, J. (1998).
Peer relations and school adjustment in ungraded primary children. Journal of
Educational Psychology, 90(3), 506-515.
Dodge, K. A. (1993). Social-cognitive mechanisms in the development of
conduct disorder and depression. Annual Review of Psychology, 44, 559-584.
East, P. L., & Rook, K. S. (1992). Compensatory patterns of support among
childrens peer relationships: A test using school friends, nonschool friends, and siblings.
Developmental Psychology, 28(1), 163-172.

Adolescent Attachment 60
Elicker, J., Englund, M., & Sroufe, L. A. (1992). Predicting peer competence and
peer relationships in childhood from early parent-child relationships. In R. Parke & G.
Ladd (Eds.), Family-peer relationships: Modes of linkage (pp. 77-106). Hillsdale, NJ:
Erlbaum.
Epstein, J., & McPartland, J. (1976). The concept and measurement of the
quality of school life. American Educational Research Journal, 13, 15-30.
Epstein, J., & McPartland, J. (1978). The Quality of School Life Scale:
Administration and Technical Manual. Boston: Houghton Mifflin.
Fagot, B. I. (1997). Attachment, parenting, and peer interactions of toddler
children. Developmental Psychology, 33(3), 489-499.
Feldman, S. S., & Wentzel, K. R. (1990). Relations among family interaction
patterns, classroom self-restraint, and academic achievement in preadolescent boys.
Journal of Educational Psychology, 82(4), 813-819.
Finnegan, R. A., Hodges, E. V. E., & Perry, D. G. (1996). Preoccupied and
avoidant coping during middle childhood. Child Development, 67(4), 1318-1328.
Freitag, M. K., Belsky, J., Grossmann, K., Grossmann, K. E., & ScheuererEnglisch, H. (1996). Continuity in parent-child relationships from infancy to middle
childhood and relations with friendship competence. Child Development, 67(4), 14371454.
George, C., Kaplan, N., & Main, M. (1996). Adult Attachment Interview.
Unpublished manuscript, Department of Psychology, University of California, Berkeley
(3d ed.).

Adolescent Attachment 61
Harter, S. (1988). Manual for the Adolescent Self-Perception Profile. Denver:
Author.
Hoge, D., Smit, E. K., & Hanson, S. L. (1990). School experiences predicting
changes in self-esteem of sixth- and seventh-grade students. Journal of Educational
Psychology, 82(1), 117-127.
Jacobsen, T., Edelstein, W., & Hofmann, V. (1994). A longitudinal study of the
relation between representations of attachment in childhood and cognitive functioning in
childhood and adolescence. Developmental Psychology, 30(1), 112-124.
Jacobsen, T. & Hofmann, V. (1997). Childrens attachment representations:
Longitudinal relations to school behavior and academic competency in middle childhood
and adolescence. Developmental Psychology, 33(4), 703-710.
Kerns, K. A., Cole, A., & Klepac, L. (1996). Peer relationships and
preadolescents perceptions of security in the child-mother relationship. Developmental
Psychology, 32(3), 457-466.
Kobak, R., & Cole, C. (1994). Attachment and metamonitoring: Implications for
adolescent autonomy and psychopathology. In D. Cicchetti (Ed.), Rochester symposium
on development and psychopathology: Vol. 5. Disorders of the self (pp. 267-297).
Rochester, NY: University of Rochester Press.
Kobak, R. R., Cole, H. E., Ferenz-Gillies, R., Fleming, W. S., & Gamble, W.
(1993). Attachment and emotion regulation during mother-teen problem solving: A
control theory analysis. Child Development, 64, 231-245.

Adolescent Attachment 62
Kobak, R. R., & Sceery, A. (1988). Attachment in late adolescence: Working
models, affect regulation, and representations of self and others. Child Development, 59,
135-146.
Ladd, G. W., Birch, S. H., & Buhs, E. S. (1999). Childrens social and scholastic
lives in kindergarten: Related spheres of influence? Child Development, 70(6), 13731400.
Ladd, G. W., Kochenderfer, B. J., & Coleman, C. C. (1996). Friendship quality
as a predictor of young childrens early school adjustment. Child Development, 67(3),
1103-1118.
Laible, D. J., Carlo, G., & Raffaelli, M. (2000). The differential relations of
parent and peer attachment to adolescent adjustment. Journal of Youth and Adolescence,
29(1), 45-59.
Lieberman, M., Doyle, A., & Markiewicz, D. (1999). Developmental patterns in
security of attachment to mother and father in late childhood and early adolescence:
Associations with peer relations. Child Development, 70(1), 202-213.
Main, M., & Goldwyn, R. (in press). Adult attachment rating and classification
systems. In M. Main (Ed.), A typology of human attachment organization assessed in
discourse, drawings and interviews (working title). New York: Cambridge University
Press.
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and
adulthood: A move to the level of representation. Monograph of the Society for Research
in Child Development, 50, 66-104.

Adolescent Attachment 63
Nada Raja, S., McGee, R., & Stanton, W. R. (1992). Perceived attachments to
parents and peers and psychological well-being in adolescence. Journal of Youth and
Adolescence, 21(4), 471-485.
Pianta, R. C., Nimetz, S. L., & Bennett, E. (1997). Mother-child relationships,
teacher-child relationships, and school outcomes in preschool and kindergarten. Early
Childhood Research Quarterly, 12, 263-280.
Repetti, R. L. (1996). The effects of perceived daily social and academic failure
experiences on school-age childrens subsequent interactions with parents. Child
Development, 67(4), 1467-1482.
Scaramella, L. V., Conger, R. D., Simons, R., & Whitbeck, L. B. (1998).
Predicting risk for pregnancy by late adolescence: A social contextual perspective.
Developmental Psychology, 34(6), 1233-1245.
Slough, N. M., & Greenberg, M. T. (1990). Five-year-olds representations of
separation from parents: Responses from the perspective of self and others. In I.
Bretherton & M.W. Watson (Eds.), New directions for child development: No. 48.
Childrens perspectives on the family (pp. 67-84). San Francisco: Jossey-Bass.
Sroufe, L. A., & Waters, E. (1977). Attachment as an organizational construct.
Child Development, 48, 1184-1199.
Ward, M. J., & Carlson, E. A. (1995). Associations among adult attachment
representations, maternal sensitivity, and infant-mother attachment in a sample of
adolescent mothers. Child Development, 66, 69-79.

Adolescent Attachment 64
Wentzel, K. R. (1994). Relations of social goal pursuit to social acceptance,
classroom behavior, and perceived social support. Journal of Educational Psychology,
86(2), 173-182.
William T. Grant Foundation Commission on Work, Family, and Citizenship.
(1988). The forgotten half: Pathways to success for Americas youth and young families.
Washington, D.C.: William T. Grant Foundation Commission on Work, Family, and
Citizenship.

Вам также может понравиться